49

SOME MALAYSIAN PHYTOGEOGRAPHICAL PROBLEMS.

B y E. D . MERRILL, Professor of Botany , Harvard University. Perhaps no part of the l\·orld is more intriguing from the standpoint of phytogeography than is the great equatorial archi­ pelago lying bet\Yeen southern Asia and Australia. :Malaysia is by far the largest island gronp in the world, lies ,,·holly \Yithin the humid tropics, has great diversity of altitudes np to nearly five thousand metres, and enjoys uniformly high low altitude temperatures, and, except in liibited regions, an abundant rainfall. Almost continuous precipitation occurs over large sections, accompanied by relatively high humidity; other large areas are characterized by alternating wet and dry seasons. These factors, combined with the characters of the soils, the topography, and the position of mountain masses in relation to prevaili ng winds provide optimum conditions for plant grm,·th, and the net result is a flora of tremendous richness and exuberance. The differentia­ tion of species has perhaps been in part favot·ed by the geoloo·ic development of the region, and its more or less insular character over long periods of time. Under present conditions contiguous parts of the same island may present rather strikingly different floras, \Yhile certain islands separated from each other by onl~· relatively narrow anns of the sea may have very strikingly different vegetative and floristic aspects. Some years ago on the basis of a study of endemism of thos~ parts of Malaysia blessed with published floras or compiled enumerations, such as the Malay Peninsula, Java, Borneo, and the Philippines, I concluded that when the flora was approxi­ mately known, that in the Pteridophytes and the Spermatophytes combined its total "·ould approximate to 45,ooo species. Thus for the Malay Peninsula, Borneo, and the Philippines, considering all kno\\·n species, it was {ound that at least fifty per cent. of those known from each major unit \Yere endemic. Considering how relatively little \\·as then, and still is, knmn1 regarding the floras of the large islands, including Snmatra, Borneo, Celebes, and New Guinea, I am inclined to consider that this estimate o£ 45,000 species is perhaps too conservative. With a land area in excess of one million square miles. this region (including the Malay Peninsula, which is almost an island, and \Thich is essentially a part of the Archipelago) extends from northern Sumatra to Java, Ne\v Guinea, and the Philippines, .-t region of magnificent distances, extending from north-east to south-east, a distance of over forty-five hundred miles. Among its many thousand islands are found the largest ones in the world : New Guinea, Sumatra, and Borneo, and such large

Vol. IX. (1935). so

secondary ones as Java, Luzon, 1\lindanao, Celebes, Gilolo, and Timor. The magnitude of some of these major islands is well suggested by Wallace's statement that 'Yithin Borneo "the ,,·hole of the British Isles might be set don·n, and n·ould be surrounded by a sea of forests." This is one of those parts of the world regarding \\"hich ''"e have only an approximate knowledge of its flora. Certain areas, such as parts of the Malay Peninsula, Java, and parts of the Philippines, may be considered to be n·ell-known botanically, ·but in fact only certain very limited parts are thoroughly kno,rn. Vast regions have been only superficially explored, and great areas have never been visited by any collector or botanist. A distinctly high percentage of the known species is known only from the single collections on which the original descriptions are based. We cannot, therefore, make any reasonably accurate com­ parison between any particular part of this vast region and comparative areas in what may consider to be thoroughly explored parts of Europe and North America. In any consideration of the numerous phytogeographical problems of Malaysia we are handi­ capped by our ovvn limited knowledge of its exuberant flora, and must realize that future intensive exploration will yield data which may conceivably alter some of our at present accepted axwms. Although the geologic history of Malaysia is reasonably ''"ell understood in its salient outlines much detailed knowledge is lacking. The magnificent work of the Dutch geologists and hydrographers on the geological history of Malaysia forms a basis that must be considered by those biologists interested in the present-day distribution of plants and animals, for the key to present-day distribution to a large degree is found in geolog-ic history. Whatever the causes may have been, it seems to be evident that the great islands of Sumatra, Java, Borneo, and the Palawan-Calamian group in the Philippines were in the Pleiocene­ Pleistocene at times connected with the Asiatic continent, and Nen· Guinea had the same relationship with Australia. There is much evidence to support the idea that most of the Philippine group, Celebes, Lesser Sunda Islands, and the Moluccas, at this time in geologic history, as "\\ell as in the Tertiary, n·ere con­ stantly archipelagic, with no direct land connections between eastern and western Malaysia, and hence no direct land connec­ tions between Asia and Australia since the early Tertiary. Botanists, probably because confronted by such myriads of species, have not been as prone as zoologists to delimit life zones or to map provinces. The attempts of the zoologists to establish lines of demarcation separating the Asiatic and Australian faunas are interesting. Wallace's line extending northward from between Bali and Lombok through the Macassar Strait between Borneo and Celebes is, at least popularly, the best known of these. Its most noteworthy substitute is Weber's Line, farther

Gardens Bulletin, S .S. 51 to the east. If these lines are plotted on a hydrographic map the former approximates the eastern boundary of the Asiatic contin­ ental shelf and the latter the western boundary of the Australian­ Papuan shelf. It is reasonably good evidence that these two shelves mark the approximate limits of previously existing continental areas. These two more or less stable areas havt been constantly separated by a tri~ngular unstable insular area, its base the Lesser Sunda Islands, its apex Luzon. There is no sharp line of demarcation between the Australian and the Asiatic faunas and floras, but rather this unstable insular area forms a great transition zone that has been in part populated by infiltrations from the south and east and from the north and ,rest. No proposed line is absolute for any single group, much less when all groups are considered. It is a well-known fact that as we progress southward and eastward from Asia, the continental types of animals and plants become fewer and fewer, and as we progress northward and eastward from A."ustralia, the Australian types likewise decrease even more markedly. This is exactly what we would expect tf our present concept of the geologic history of the region is approximately correct. One should not make the mistake of drawing final conclu­ sions on the basis of the present-day distribution of any one group of organisms, whether it be the molluscs, insects, reptiles, birds, freshwater fishes, mammals, or plants, and much less on any single family or genus within these major groups; and yet there are conspicuous cases of broad generalizations that have been founded on the study of a single family, perhaps even a single genus. While some such generalizations may ultimately provt> to be correct, it is always safer to consider what is known regard­ ing the geographic distribution of other and perhaps totally unrelated groups and to attempt to correlate such data vvith what is known of the geologic history of the entire region. It is a safe assumption that the different groups have been developed and have attained their present-day distribution at different periods in geologic history; and that while some groups were rapidly differentiating, others were perhaps static or were retrograding. That the Philippine flora, as to the Myrtacec:e, shows an un· mistakably close relationship with that of north-eastern Australia is no reason for considering that the entire Philippine flora and fauna is allied to that of Australia. If one studies only the Dipterocarpacere of the Philippines one would conclude that its flora is overwhelmingly allied to that of Borneo, Sumatra, and the Malay Peninsula. Clearly here, as with other groups, the periods of development and expansion have occurred at different times in geologic history, periods that permitted the distribution 0£ one group perhaps restricting that of another.

Vol. IX. (1935). 52

Botani_cal collectors who have explored Malaysia compared \Yith those who have \Yorked in Europe and in Xorth America, north of Mexico, have been fe,,:, and the number of botanists who have familiarized themselves " ·ith \\·hat has been published is fewer still. For all practical purposes those botanists \\·ho have concentrated on a study of the floras of the :\Ialay Peninsula, Sumatra, Java, Borneo, the Philippines, and ~e\\" Guinea, in the past t\YO hundred years are merely the pioneers, and their \York should be considered as that of pioneers, to be correlated, polished, and evaluated by future generations of botanists who \Vill have infinitely more to work with than did those ,,·ho first exploited the fertile fields of Malaysian botany. It "·ill be many years before the essential data are assembled on \Yhich \Ye may base final conclusions. Even such work as has been done by various taxonomists on the Malaysian flora \\'Orking in scattered centres in Europe, England, and America, in Manila, Singapore, Buiten­ zorg, and Calcutta, has not been \\"ell correlated. :M:uch needs to be done to prove the validity or non-validity of some thousands of proposed species. Many points can be settled only \Yhen competent taxonomists, having access to comprehensive collec­ tions, shall have harl the time and opportunity to prepare and publish authoritative revisions of numerous natural groups. As this is done, and as type collections and other authentic material are critically studied and the species carefully compared \\·ith each other, nomenclatural changes and many reductions of pro­ posed species will be the order of the day. In accordance with the current aims and methods of botanical \\"Ork such changes \Yill be inevitable. In the more thoroughly explored parts :)f the \\·orld, and in those countries where their floras have been long and intensively studied, we approach a condition of finality but in such vast regions as Malaysia, \Yhere great areas still remain to be explored and even existing collections have not been intensively and comparatively studied, our taxonomic con­ clusions cannot be other than approximate. It has long been known that \\·ithin Malaysia \\·e find ''"hat ... we consider to be Asiatic and Australian types. To a much lesser degree we find scattered African and Mascarene types from the west, Polynesian ones from the east, and N e\T Zealand and New Caledonian ones from the southeast. Both the Asiatic and Australian types include high and low altitude forms. In refer­ ence to the high altitude and some of the low altitude Asiatic types '"-e are reasonably confident that they actually are continental in origin, because in such genera as Taxus, Pinus, Ulmtts, Salix, Acer, Prirnula, Gledif,sia, Lilium, Anemone, Sedu.m, Rosa, Bux us, Potentilla, L·ysimachia, Aster, Solidago, anrl many others, the numerous species are characteristic of the north temperate zone of both hemispheres. The occurrence of scattered species in Malaysia is unquestionably due to the fact that somehow at some

Gardens Bulletin, S.S. 53 time in geologic history these representatives of northern genera extended their ranges southward into Malaysia $lnd have persisted there. As to the Australian types in :Malaysia such as Eucalyptus, Centrolepis, Patersonia, Drim.ys, Clianthus, Stackhousia, Pimelea, Tristania, LeptosPeTmum, Melaleuca, Xanthostem.on, Halorhagis, Didiscus, St·y lidium, and others, we can also be reasonably certain that they are true Australian types because of the great develop­ ment of these genera in Australia and , to a certain degree, in New Caledonia; and that the fe\Y representations in Malaysia mio-rated north\Yard into the archipelago even as the Asiatic types migrated south\Yard into J'vialaysia. There are, ho\\·ever, a very large number of genera common to north-eastern Australia and Malaysia and southern Asia and Malaysia, some covering the entire range from Asia to Australia, whose origins it is practically impossible to determine and to assign to Asia rather than to Malaysia or -z.1ice versa. Doubtless many of these were originally Malaysian and extended their ranges southeastward to Australia and northward to Asia, but many others probably Asiatic. Their universal distribution within the monsoon region is unquestionably in part a reflex of the adapta­ bility of these plants to a certain type of climatic conditions. There are relatively few African and Mascarene types in Malaysia, such as AngrcecU'm., Combretodendron (Petersia, Pefersianth:us), E1·ythrophlceum, and Baphia, but definitely they are African and Mascarene because the principal development of species in these genera (at least Erythrophlreum, A ngrcecum. and Baphia) is found in those regions. When \Ve come to the so-called Polynesian types in Malaysia, such as Vavcea, I oinvillea, Dolicholobum, Tetraplasandra, and Cou.thovia, the present indications are that they are not really Polynesian types in Malaysia but are actually Malaysian types in Polynesia, the generic types merely having been first described from the latter region. Thus Vavcea is now known to occur from Java, Borneo, the Philippines, and NeiY Guinea; and New Guinea in definitely the center of origin and dispersal of Couthonia and of Dolicholo­ biunt. I oinvillea .and T etraplasa.ndra, both small genera, have Malaysian representatives. The numerous definitely Asiatic types in Malaysia apparently followed certain distinct routes of migration into Malaysia from the continent. In the iYest unquestionably the most important route was through the Malay Peninsula to Sumatra, Borneo, Java, and even further to the south and east. Considerable evidence leads us to believe in the existence of a secondary western route from Asia through the Andaman and Nicobar Islands to Sumatra. In the east, considering the fact that certain Asiatic types occur in the Philippines and in Celebes, but are absent from other parts of Malaysia, there i\"as apparently another route from southeastern Asia to Formosa acd the Philippines and

Vol. I X . (1935). 54 southward. We find in Formosa and on the higher mountains of northern Luzon certain Himalayan types, in many cases these representing the most eastern and southern extensions of the Himalayan flora. Increasing knowledge of the biology of Malaysia is tending more and more to confirm tlie belief that there have been two great centres of origin and dispersal of tropical types, one in the west, the Sunda Islands, Java, Borneo, Sumatra, and the Malay Peninsula, and one in the east, New Guinea. From the latter plants and animals migrated southward into eastern Australia, eastward into Polynesia, and northward into the Philippines, few extending westward to the Sunda Islands From the Sunda Islands the striking migrations were northward into tropical Asia, northeastward into the Philippines, and to a more limited degree eastward and south·ward in the direction of New Guinea. There were, of course, intermigrations within each area. Comparative studies of the climates and plants in tropical Asia and Malaysia with those of America, emphasizes one striking difference. In America the humid tropics are to a large degree separated from the temperate regions by ex tensive arid areas. In the Old World the intervenin~ arid regions are lacking . There has been a continuous humid north and south route connecting Malaysia with subtemperate and temperate parts of Asia-the Malay Peninsula northvvard through Burma, Siam and Indo­ China, to India and southern and eastern China. The limiting f~ctor here has been largely temperature, while in America it has been aridity and temperature combined. We have little evidence on which to base conclusions as to what the humidity and aridity of the various parts of Malaysia may have been in past geologic times, but the available paleobotanical evidence seems to indicate continuously humid tropics in the past, in general similar to the conditions existing to-day. Outside of the taxonomic field, but nevertheless intimately associated with and depend~nt upon it, there are an infinite number of problems to be settled regarding the Malaysian fauna and flora. In part these depend on an intensive study of existing collections of reference material, in part on long continued inten­ sive field work. What was the original flora of Malaysia? What are to be considered the relic species of earlier floras ? What factors permitted or encouraged the inter-migration of Australian­ Malaysian and Asiatic-Malaysian types and ';vhat were the limiting factors? Were these geological or climatic in nature, or both? Does the theory of continental shift throw any light on the problem ? What are the relations between the present distribution of plants and the prevailing winds? What bearing do typhoons have on the problem? What is the relationship of existing ocean currents to the spread of plants? What does the annual north and south flight of certain types of mig-ratory birds signify for those plants characterized by the production

Gardens Bulletin, S .S. 55

of myriads of minute seeds and adapted to muddy or wet habitats? What are the e-ffects of constantly moist and alter­ natingly wet and dry conditions? \Vhat is the e-ffect of dominant mountains and mountain ranges on prevailing "\Yinds and vice versa, and how do these factors affect vegetation? What is the bearing of geologic history and of hydrography and topography on the problem ? From long residence in the tropics and from numerous observations made on the vegetation, some years ago I reached the general conclusion that in modern times man is the most important single factor in the dissemination of plants. Manifestly much of Malaysia was originally covered by a forest vegetation of one type or another. To-day vast areas are under intensive cultivation, almost equally large areas are in primitive cultiva­ tion, enormous tracts are covered with coarse grasses, bamboo thickets, and second growth forests. Man's agricultural activities have been the greatest single factor in the destruction of the original forests over vast areas, but lumbering, and in some regions mining , has played a part. It is axiomatic that in Malaysia a primary forest once destroyed is never re-established directly; its place is taken by coarse grasses, bamboo thickets, or dense stands of rapidly growing small trees and shrubs. As the primary vegetation has been destroyed the habitats have been provided for plants from other regions. These, accidentally or intentionally introduced by man, or through the agency of nature, once man has paved the way for them, have become dominant over great areas. Many of these were of American origin, introduced since the sixteenth century, including representation • of such woody genera as Leu..c cena, Psidium, Lantana, Pithe.cello­ bium and ProsoPis, and a very large number of weedy herbaceous plants in such genera as D?"ym.aria, M alachra, W altheria, Stachytarpheta, Cassia, Hyptis, Agemtum., Cosm,os, Elephan­ topus, Erechtites, and Synedrella, as well as numerous grasses and sedges. Antedating these numerous American introductions were many Asiatic types, and this introduction of Old World forms is still progressing. There is much that needs to be done in reference to the history of these introduced plants. Where did they come from? When were they introduced, and how ? What are the reasons for their present dominancy? What is the percentage of introduced species in primary forest areas as com­ pared with the percentage in those regions vvhere man is directly or indirectly responsible for the present vegetation? Endemism presents numerous fascinating problen1s. Throughout Malaysia one finds in most of the large genera a few strongly characterized, widely distributed species-. At the same time in these genera one notes that in such groups as Ficus, Antidesma, Eugenia, Medinilla, Dendrobium, Quercus, Eria, Bulbophyllum, Cala1nus, Aglaia, B egonia, Polyalthia, and Canarit£m, to mention only a few, that there is a tremendously

Vol. I X. ( 1935). s6 developed local endemism in many of the larger islands and even in some of the smaller ones. Invariably the percentage of endemics is very Ion· in the settled areas, ope grass lands, and secondary forests, and correspondingly is very high in those groups that require primary forest conditions. One concludes that the primary forests represent the original vegetation, and that the Ion· percentage of endemics in the settled areas, open grass lands, and secondary forests is a reflection of the activities of man on the vegetation as a ·whole; that many of the species no'iY dominant in these latter areas were introduced as the primary forests \Yere destroyed. We are reasonably certain that this 'i\·as the case in reference to the many species of American origin, "·hich predominate in the settled areas, for these did not commence to come in until some time during the sixteenth century. We are less certain in reference to those species of Asiatic origin, largely groups characteristic of the second growth forests and open grass lands. The cultivated plant s represent a series of problems in themselves. l\Iost of those in Malaysia are not natives of the archipelago but \\"ere introduced by man from other regions . T he earliest invading peoples probably brought no plants for they had no agriculture. Later ·waves of more advanced peoples cer­ tainly brought cultivated plants ·with them. Many basic Malaysian food plants were introduced in prehistoric times, others in modern times; and the introduction of 'i'i"eeds parallels the introduction of economic species. Within the period of western expansion the Malaysian flora has been greatly enriched through the introduction of numerous economic plants and many weeds • of American origin. Not to be disassociated from the cultivated plants are their aboriginal names, "·hich in many cases point definitely to the place of origin and the approximate time of introduction of individual species. In these fields of ethnobotany and compara­ tive philology of plant names much needs to be done, and Malaysia n·ith its very numerous dialects and its great racial mixtures offers a marvellously fertile field. Again in the field of plant biology there are numerous problems. The relationships between insects and plants, birds and plants, · and even bats and plants in reference to cross pollinatiou. The biological relationships between ants and such domiciliary plants as H oya, Dischidia, M·yrnteconauclea, Neona·uclea, .11acaranga, L ecanopteris, Polypodium, Dcemonorops, Myrmecodia, H~vdnophytulll, Mynnedom,a, Saurauia, and Piper, and representatives of other genera where plant organs such as branches, stems, leaves, and sheaths have been modified to form domiciles for colonies of ants. Rut this and numerous other subjects is outside the scope of this paper.

Gardens Bu.lletin, S .S. 57

In this brief survey, hurriedly written because of the exigencies of my present situation, and entirely undocumented, I have attempted merely to ~numerate some of the numerous intriguing problems of Malaysian phytogeography. I have done so in the hope that some "·ho read it may be tempted to assist in solving some of the problems by undertaking field work in this too little explored part of the world or to co-operate in the study of the rich reference collections already available. In Malaysia one is constantly on the edge of the botanical unknown. The field is vast, existing problems create further ones, and the need of more investigators is great. The serious-minded interested student ''"ho elects to assist in the study of the plants and plant problems of Malaysia will be amply repaid by the productive results secured.

Vol. IX. (1935).